RG posted nefs from the D800 and D800E, best direct comparison shots we can process from raw up that I've seen so far: Rob Galbraith DPI: Comparing detail and moire in the Nikon D800 and D800E
I downloaded the first set (image above) with the shoes and motorcycles and been playing with them in C1. He used the excellent 85/1.4 at optimal f5.6 aperture for both captures, conditions remained almost the same for both.
I opened them in C1, set sharpening on them the same and no surprise the E looks a little better. I then experimented to find a more ideal setting for each and ended up setting the E at 120/0.6/2.0 which appears to be right at the edge of being oversharp. I then adjusted settings on the regular file to a similar look, which took sharpening up to 240/0.6/2.0 and required a clarity bump to 25 to get teh micro contrast similar -- and to my eyes they look almost identical at actual pixel view. I left everything else at base C1 settings. There is perhaps a VERY slight edge to the E except it has the pattern and color moire in the shoes and left-most black motorcycle engine center frame. Next I slide the moire tool up to where it dissipates partially and matches the D800 -- at about 48 amount and 70 pattern -- they are for all intents and purposes dead nuts equal looking files onscreen at 100% view, but with a very slightly smoother "less digital" look to the regular D800 file. Note too, that the regular D800 file retains some of the pattern moire, but with much lower color. So far fine, at actual pixel view I can see some differences, but from experience I believe them to be extremely minor for normal uses. Now it was time for the ultimate comparison, a print, and so print I did.
I loaded the C1 tiffs into CS, applied my basic output sharpening (very, very light) and printed them each at native size on my Epson, to 13-1/2" x 20-1/2" at 360PPI on Harman FBAL Gloss Baryta, a very sharp paper. I left the moire tool off in C1 so the D800E file would get its full resolution benefit, but obviously would still show the moire. Viewing the prints with my reading cheaters on and nose in the prints, I can honestly see no difference -- and am a little surprised because I can't even really detect the moire in the E file as it is such a small area of the print. So I get the loupe out. With the loupe, I can see the moire more clearly and thus tell which print is which, but it honestly took the loupe for my eyes. Interesting point here is under the loupe, I can read the miniscule print in the blue no skateboarding sign in the D800 print to the right of the shoes next to the bench, but CANNOT read it in the D800E print due to moire color pollution occluding the type!
So next I uprez both to 2x native size at 27" x 41" and print comparative sections. First thing to comment on is both cams hold that size fine, but are starting to show some haggered edges -- so they won't go much larger for nose in the print viewing, but my guess is still fine for a 40" x 60" print at normal viewing distances. Anyway, with my reading cheaters on and nose in the prints, they look identical. I get the loupe out, they still look almost identical, and I am having a hard time choosing which is which because the moire is attenuated slightly by the uprez. (I need to investigate this more, not sure why or how, but it was a real result, probably due to the "smoothing" in the uprez algorithm.) Anyway, about now I am wondering if I printed the same image twice by mistake, so I reprint them and confirm they are indeed that close.
In conclusion for me so far, it's a REALLY tough call to say the E is "better" assuming you process them both IDEALLY instead of IDENTICALLY. I personally think the differences are academic for print, and remain so unless you plan on showing 100% crops on a large monitor as your main viewing platform. My study continues, but for right now I'm giving the regular D800 the slight advantage for MY uses simply because of the lesser moire issues. YMMV…
Special thanks to Rob Galbraith for making these files available!
Cheers,
I downloaded the first set (image above) with the shoes and motorcycles and been playing with them in C1. He used the excellent 85/1.4 at optimal f5.6 aperture for both captures, conditions remained almost the same for both.
I opened them in C1, set sharpening on them the same and no surprise the E looks a little better. I then experimented to find a more ideal setting for each and ended up setting the E at 120/0.6/2.0 which appears to be right at the edge of being oversharp. I then adjusted settings on the regular file to a similar look, which took sharpening up to 240/0.6/2.0 and required a clarity bump to 25 to get teh micro contrast similar -- and to my eyes they look almost identical at actual pixel view. I left everything else at base C1 settings. There is perhaps a VERY slight edge to the E except it has the pattern and color moire in the shoes and left-most black motorcycle engine center frame. Next I slide the moire tool up to where it dissipates partially and matches the D800 -- at about 48 amount and 70 pattern -- they are for all intents and purposes dead nuts equal looking files onscreen at 100% view, but with a very slightly smoother "less digital" look to the regular D800 file. Note too, that the regular D800 file retains some of the pattern moire, but with much lower color. So far fine, at actual pixel view I can see some differences, but from experience I believe them to be extremely minor for normal uses. Now it was time for the ultimate comparison, a print, and so print I did.
I loaded the C1 tiffs into CS, applied my basic output sharpening (very, very light) and printed them each at native size on my Epson, to 13-1/2" x 20-1/2" at 360PPI on Harman FBAL Gloss Baryta, a very sharp paper. I left the moire tool off in C1 so the D800E file would get its full resolution benefit, but obviously would still show the moire. Viewing the prints with my reading cheaters on and nose in the prints, I can honestly see no difference -- and am a little surprised because I can't even really detect the moire in the E file as it is such a small area of the print. So I get the loupe out. With the loupe, I can see the moire more clearly and thus tell which print is which, but it honestly took the loupe for my eyes. Interesting point here is under the loupe, I can read the miniscule print in the blue no skateboarding sign in the D800 print to the right of the shoes next to the bench, but CANNOT read it in the D800E print due to moire color pollution occluding the type!
So next I uprez both to 2x native size at 27" x 41" and print comparative sections. First thing to comment on is both cams hold that size fine, but are starting to show some haggered edges -- so they won't go much larger for nose in the print viewing, but my guess is still fine for a 40" x 60" print at normal viewing distances. Anyway, with my reading cheaters on and nose in the prints, they look identical. I get the loupe out, they still look almost identical, and I am having a hard time choosing which is which because the moire is attenuated slightly by the uprez. (I need to investigate this more, not sure why or how, but it was a real result, probably due to the "smoothing" in the uprez algorithm.) Anyway, about now I am wondering if I printed the same image twice by mistake, so I reprint them and confirm they are indeed that close.
In conclusion for me so far, it's a REALLY tough call to say the E is "better" assuming you process them both IDEALLY instead of IDENTICALLY. I personally think the differences are academic for print, and remain so unless you plan on showing 100% crops on a large monitor as your main viewing platform. My study continues, but for right now I'm giving the regular D800 the slight advantage for MY uses simply because of the lesser moire issues. YMMV…
Special thanks to Rob Galbraith for making these files available!
Cheers,